Knowledge consumption, retrieval and retention call for different structures of information

Key idea: we might need to trans­form infor­ma­tion between dif­fer­ent struc­tures at dif­fer­ent stages of learning:

When we build a cor­po­rate knowl­edge base, we usu­ally think of it as a book or a col­lec­tion of instructions. That way, we may unknow­ingly put read­ing it for the first time ahead of using it for ref­er­ence later.

Here’s how employ­ees actu­ally inter­act with information.

Consumption (learn­ing). Sequential struc­tures work better at first, because they align with our cog­ni­tive limits and our nat­ural learn­ing process. Humans nat­u­rally under­stand and recall infor­ma­tion in a sequen­tial, chrono­log­i­cal or cause-and-effect order. Sequentially struc­tured infor­ma­tion is also easier to transmit. Reading is decom­press­ing a nar­ra­tive into a con­cept net­work, and that decom­pres­sion is part of how we first con­sume and process information.

Writing com­presses a con­cept net­work into a nar­ra­tive; read­ing decom­presses it

Retrieval. Network struc­tures are supe­rior here, because they allow mul­ti­ple access points and path­ways to the same infor­ma­tion: a net­worked narrative. A net­work makes infor­ma­tion easy to reach, which is cru­cial for PKM, while ease of trans­mis­sion doesn’t matter in this context. This aligns with how our brains actu­ally store infor­ma­tion — in inter­con­nected seman­tic net­works, not linear sequences. Multiple path­ways mean we can retrieve infor­ma­tion even if we forget the exact “route” by which we orig­i­nally learned it.

From Ilya Shabanov’s The 10 Commandments of Academic Note-Taking

Retention. A hybrid approach seems most effective. While we might learn sequen­tially, reten­tion works best when infor­ma­tion is inte­grated into a broader net­work of understanding. Pattern recog­ni­tion (network-based) and sequence recog­ni­tion (linear) serve dif­fer­ent but com­ple­men­tary pur­poses in retention.

We should build knowl­edge man­age­ment tools with these processes in mind.

A corporate knowledge base that respects both structures

Entry points. Create mul­ti­ple “front doors” for dif­fer­ent use cases: task-based path­ways (“I need to do X”), concept-based nav­i­ga­tion (“I want to under­stand Y”), role-based guides (“I’m a new sales rep”) and problem-based access (“I’m trying to solve Z”).

Sequential learn­ing paths. Design struc­tured learn­ing sequences for first-time exposure. Break com­plex topics into digestible mod­ules, pro­vide clear pro­gres­sion paths, and include check­points and prac­ti­cal exercises. Example: onboard­ing guides that follow a “day 1 → week 1 → month 1” structure.

Network connections. Implement rich hyper­link­ing between related con­cepts, create con­cept maps that show rela­tion­ships, use con­sis­tent tag­ging for cross-referencing, and build “see also” sec­tions that show related content. Example: each pro­ce­dure links to related pro­ce­dures, the con­cepts beneath it and common troubleshooting.

Retrieval features. Strong search that han­dles task-based queries, quick ref­er­ence guides and check­lists, context-sensitive help inside tools, and visual nav­i­ga­tion maps. Example: search that under­stands both “how to process refund” and “refund policy”.

Retention support. Practice sce­nar­ios and exam­ples, real-world case stud­ies, deci­sion trees for com­plex processes, and inter­ac­tive ele­ments for active engagement. Example: trou­bleshoot­ing guides that com­bine step-by-step processes with an under­stand­ing of the under­ly­ing systems.

Integration. Link sequen­tial guides to network-based ref­er­ence mate­r­ial, pro­vide both quick-start guides and deep dives, create mul­ti­ple paths to the same infor­ma­tion, and build bridges between the­o­ret­i­cal and prac­ti­cal content. Example: a pro­ce­dure guide that links to both “next steps” and related concepts.